As I want to have my astrometry in Alt Az at the end, I wanted to check how refraction affects solving. I have a 10x15 deg photo with center in ~18deg Alt (so lowest part of photo should be on ~13 deg). Accourding to wiki refraction on ~13 deg Alt should be ~5arcmins, but after solving the image with --tweak-order 1 there is no trace of refraction. I'm checking this by comparing field_x,y with index_x,y, with pixel scale of my photo ~10arsecs/pix I was expecting around 10-15 pix residuals in lower part of my photo, but there are only ~2 pix errors. How is that refraction isn't visible in those resiudals?